CSS software tools sought

[andrew]

Re: CSS software tools sought

On 2007-06-17, Albert Wiersch <nospam@nospam. nospamwrote:

[....]

What is more important is that things get done in the best way that is
reasonable, and often that means forgetting about complete technical
conformance, which, in many cases, is a waste of time when it doesn't bring
any practical benefit.

Now that is a slightly scary statement. Does the expression "Near
enough is good enough" effectively paraphrase this?

Andrew

--
Andrew's Corner

http://people.aapt.net.au/~adjlstrong/homer.html

[dorayme]

Re: CSS software tools sought

In article <f52o1f$ffl$1@n ews-01.bur.connect. com.au>,
andrew <nospam@andrew. invalidwrote:

On 2007-06-17, Albert Wiersch <nospam@nospam. nospamwrote:
>
[....]
>

What is more important is that things get done in the best way that is
reasonable, and often that means forgetting about complete technical
conformance, which, in many cases, is a waste of time when it doesn't bring
any practical benefit.

>
Now that is a slightly scary statement. Does the expression "Near
enough is good enough" effectively paraphrase this?

What AW says in this quote is hardly scary even given your
paraphrase. After all, it is applied in this newsgroup to all
manner of things, and rightly so. It is a simple nonsense to
suppose that a lack of perfection in something is necessarily a
bad or scary thing. I assume that JK and others are getting stuck
into this guy because people can actually fail to get practical
benefits...

--
dorayme

[Ben C]

Re: CSS software tools sought

On 2007-06-17, dorayme <doraymeRidThis @optusnet.com.a uwrote:

In article <f52o1f$ffl$1@n ews-01.bur.connect. com.au>,
andrew <nospam@andrew. invalidwrote:
>

>On 2007-06-17, Albert Wiersch <nospam@nospam. nospamwrote:
>>
>[....]
>>

What is more important is that things get done in the best way that is
reasonable, and often that means forgetting about complete technical
conformance, which, in many cases, is a waste of time when it doesn't bring
any practical benefit.

>>
>Now that is a slightly scary statement. Does the expression "Near
>enough is good enough" effectively paraphrase this?

>
What AW says in this quote is hardly scary even given your
paraphrase. After all, it is applied in this newsgroup to all
manner of things, and rightly so. It is a simple nonsense to
suppose that a lack of perfection in something is necessarily a
bad or scary thing. I assume that JK and others are getting stuck
into this guy because people can actually fail to get practical
benefits...

You are right that a line has to be drawn but the question is where. It
seems surprising to me that an HTML lint tool should not be at least a
proper validator plus some extra helpful warnings. Most tools of that
kind have "warning levels" to adjust the nannyishness. You'd think HTML
validity would be the most basic level.

Although AW's tool probably is quite useful to some people there are a
lot of "bogosity indicators": it has a misleading name; it's horrible
non-free shrink-wrapped Windows-only software with a GUI where stdout
and stderr would have served better; and Chris Johnson very quickly
found a bug in its HTML parser.

[dorayme]

Re: CSS software tools sought

In article <slrnf79vp1.9do .spamspam@bowse r.marioworld>,
Ben C <spamspam@spam. eggswrote:

On 2007-06-17, dorayme <doraymeRidThis @optusnet.com.a uwrote:

In article <f52o1f$ffl$1@n ews-01.bur.connect. com.au>,
andrew <nospam@andrew. invalidwrote:

On 2007-06-17, Albert Wiersch <nospam@nospam. nospamwrote:
>
[....]
>
What is more important is that things get done in the best way that is
reasonable, and often that means forgetting about complete technical
conformance,
>
Now that is a slightly scary statement. Does the expression "Near
enough is good enough" effectively paraphrase this?

What AW says in this quote is hardly scary even given your
paraphrase.
...

>
... You'd think HTML validity would be the most basic level.

You would think so. I agree. But how is it really to be judged?
Forget the commercial aspects, the claims for it, the
endorsements. A real question is does it help people make better
websites than they would without it? Perhaps these are the people
who are put off by the starker pronouncements from the strict
validators? It is very hard to judge.

Most of us around here kind of like _strict_. But most of the
internet is anything but. There seems to me a place for
entrepreneurial offerings that make this world a bit better, even
if opportunities are missed to make it even better.

--
dorayme

[Albert Wiersch]

Re: CSS software tools sought


"Chris F.A. Johnson" <cfajohnson@gma il.comwrote in message
news:3ilek4-h7v.ln1@206-248-139-163.dsl.teksavv y.com...

>
<http://cfaj.freeshell. org/ttc/>:
>

http://validator.w3.org/:

Valid HTML 4.01 Strict
>

http://www.htmlvalidator.com/:

Errors reported:
The end tag for "dd" (opened in line 74) should appear
before the end tag for "dl" (nesting error).
>
The HTML 4.01 specification says:
Start tag: required, End tag: optional

Yes, one advantage to CSE HTML Validator is that it enfoces better
structure, requiring some tags that are technically optional.

Albert

[Albert Wiersch]

Re: CSS software tools sought


"Chris F.A. Johnson" <cfajohnson@gma il.comwrote in message
news:1onek4-qfv.ln1@206-248-139-163.dsl.teksavv y.com...

>
Do you have a version I can run? I don't do Windows.

Sorry, we only have a Windows version. We do have an online version based on
the lite edition here:

http://onlinewebcheck.com/

But the lite edition is not as thorough and doesn't check spelling,
accessibility, CSS, and links.

Albert

[Albert Wiersch]

Re: CSS software tools sought


"Ben C" <spamspam@spam. eggswrote in message
news:slrnf79vp1 .9do.spamspam@b owser.marioworl d...

>
You are right that a line has to be drawn but the question is where. It
seems surprising to me that an HTML lint tool should not be at least a
proper validator plus some extra helpful warnings. Most tools of that
kind have "warning levels" to adjust the nannyishness. You'd think HTML
validity would be the most basic level.

Technical "validity" is not that great a thing to achieve in practicality.
CSE HTML Validator checks for many important things (many things that DTD
validators miss), is greatly configurable (like different "warning levels"),
and you can even use the DTD validator if you care to be strictly correct,
but hardly anyone does based on the feedback and questions I get about it.

Although AW's tool probably is quite useful to some people there are a
lot of "bogosity indicators": it has a misleading name; it's horrible
non-free shrink-wrapped Windows-only software with a GUI where stdout
and stderr would have served better; and Chris Johnson very quickly
found a bug in its HTML parser.

So it's so "horrible" because why exactly? You don't like the GUI? Did you
know it comes with a command line processor? I think you are talking about
things you don't know about.

As for the "parser bug", it's not a bug. It's designed to help people write
better HTML by closing many of their optional tags.

Albert

[Albert Wiersch]

Re: CSS software tools sought


"dorayme" <doraymeRidThis @optusnet.com.a uwrote in message
news:doraymeRid This-F09D27.19052117 062007@news-vip.optusnet.co m.au...

>
What AW says in this quote is hardly scary even given your
paraphrase. After all, it is applied in this newsgroup to all
manner of things, and rightly so. It is a simple nonsense to
suppose that a lack of perfection in something is necessarily a
bad or scary thing. I assume that JK and others are getting stuck
into this guy because people can actually fail to get practical
benefits...

I agree... I doubt the posters wanting a perfect and "perfect" HTML
documents ("perfect" being only that it validates in a DTD validator) world
always conform 100% to driving laws when driving by having never gone over
the speed limit and by never doing a rolling "stop" where a full stop is
technically required.

And when cooking, is it important to follow the instructions 100%, making
sure that it cooks or bakes for the exact amount of time specified, and that
all the ingrediants are measured 100% accurately? I don't think it matters
much. In many cases, recipes can be improved by changing the recipe to make
it better.

Have the posters who want "perfect" HTML documents always spell checked and
grammar checked their documents for "conformanc e" before posting them? I
doubt it. I bet they would agree that it would be practical to spell and
grammar check every single message the send or post.

Albert

[Ben C]

Re: CSS software tools sought

On 2007-06-17, Albert Wiersch <nospam@nospam. nospamwrote:

>
"Ben C" <spamspam@spam. eggswrote in message
news:slrnf79vp1 .9do.spamspam@b owser.marioworl d...

>>
>You are right that a line has to be drawn but the question is where. It
>seems surprising to me that an HTML lint tool should not be at least a
>proper validator plus some extra helpful warnings. Most tools of that
>kind have "warning levels" to adjust the nannyishness. You'd think HTML
>validity would be the most basic level.

>
Technical "validity" is not that great a thing to achieve in practicality.
CSE HTML Validator checks for many important things (many things that DTD
validators miss), is greatly configurable (like different "warning levels"),
and you can even use the DTD validator if you care to be strictly correct,
but hardly anyone does based on the feedback and questions I get about it.
>

>Although AW's tool probably is quite useful to some people there are a
>lot of "bogosity indicators": it has a misleading name; it's horrible
>non-free shrink-wrapped Windows-only software with a GUI where stdout
>and stderr would have served better; and Chris Johnson very quickly
>found a bug in its HTML parser.

>
So it's so "horrible" because why exactly?

Source code is not provided, documentation wastes far too much space
telling you "it's great" rather than what it actually does, although the
"Program Limitations" page is quite revealing. It's full of stupid
features like the "Tools", the "most powerful" of which, the template
tool, does the job of four lines of Perl or something similar.

Now I don't say this product is completely useless to everyone. I tried
it on a couple of pages and the warning messages it produced were fairly
user-friendly and informative (although no more so than those produced
by tidy) and I even think the tools might be quite useful to a rank
novice. But I'd rather novices learned about DTDs and real validators,
and how to use decent editors and how to think for themselves, because
none of that is really all that difficult if it's explained well, than
to get sidetracked by products like this.

You don't like the GUI?

No.

Did you know it comes with a command line processor?

No.

I think you are talking about things you don't know about.

I know a bogosity indicator when I see one.

As for the "parser bug", it's not a bug. It's designed to help people
write better HTML by closing many of their optional tags.

Rubbish. If that were case it would have said "<dt>: warning end tag is
optional and you've left it off: I hope you know what you're doing" or
words to that effect.

[Jukka K. Korpela]

Re: CSS software tools sought

Scripsit Albert Wiersch:

> http://www.htmlvalidator.com/:
> Errors reported:
> The end tag for "dd" (opened in line 74) should appear
> before the end tag for "dl" (nesting error).
>>
> The HTML 4.01 specification says:
> Start tag: required, End tag: optional

>
Yes, one advantage to CSE HTML Validator is that it enfoces better
structure, requiring some tags that are technically optional.

It claims that there is an _error_ when there is no violation of a
specification. It even claims that it is a nesting error, apparently because
the person who wrote the "CSE HTML Validator" has little understanding of
markup.

Thus, this is an example of its bogosity, not usefulness.

(Anyone who wants to make </ddobligatory can use a DTD that says so - and
use a validator, not a product that is incorrectly sold as a validator and
gives wrong reports.)

--
Jukka K. Korpela ("Yucca")

http://www.cs.tut.fi/~jkorpela/

[Chris F.A. Johnson]

Re: CSS software tools sought

On 2007-06-17, Albert Wiersch wrote:

>
"Chris F.A. Johnson" <cfajohnson@gma il.comwrote in message
news:3ilek4-h7v.ln1@206-248-139-163.dsl.teksavv y.com...

>>
> <http://cfaj.freeshell. org/ttc/>:
>>
> http://validator.w3.org/:
> Valid HTML 4.01 Strict
>>
> http://www.htmlvalidator.com/:
> Errors reported:
> The end tag for "dd" (opened in line 74) should appear
> before the end tag for "dl" (nesting error).
>>
> The HTML 4.01 specification says:
> Start tag: required, End tag: optional

>
Yes, one advantage to CSE HTML Validator is that it enfoces better
structure, requiring some tags that are technically optional.

In other words, it is not a validator, since there is nothing wrong
with omitting the closing tags. It is a style checker.

--
Chris F.A. Johnson <http://cfaj.freeshell. org>
========= Do not reply to the From: address; use Reply-To: ========
Author:
Shell Scripting Recipes: A Problem-Solution Approach (2005, Apress)

[Albert Wiersch]

Re: CSS software tools sought


"Jukka K. Korpela" <jkorpela@cs.tu t.fiwrote in message news:H3fdi.1797 17

>
It claims that there is an _error_ when there is no violation of a
specification.

Yes, it is a violation of CSE HTML Validator's specification.

Like I said before, if you just want to validate per the DTD and be limited
to it, then don't use CSE HTML Validator's validator engine which goes
beyond the DTD and performs many more checks. Use the included DTD validator
or some other free DTD validator.

It even claims that it is a nesting error, apparently because the person
who wrote the "CSE HTML Validator" has little understanding of markup.

Sorry, just more bashing.

(Anyone who wants to make </ddobligatory can use a DTD that says so -
and use a validator, not a product that is incorrectly sold as a validator
and gives wrong reports.)

Sure, one can mess with creating their own DTD for this issue, but why mess
with custom DTDs when one can use CSE HTML Validator and get not only this,
but many more checks which can't be expressed in a DTD? But go ahead and
just use the DTD validators as that's all you seem to care about and leave
the other tools to the people who actually understand the benefit from them.

Albert

[Albert Wiersch]

Re: CSS software tools sought


"Ben C" <spamspam@spam. eggswrote in message
news:slrnf7al47 .klt.spamspam@b owser.marioworl d...

>>
>So it's so "horrible" because why exactly?

>
Source code is not provided

I don't see that as horrible but of course you're entitled to your opinion.

documentation wastes far too much space
telling you "it's great" rather than what it actually does

You must have only read one or two pages because most of the documentation
explains how to use the program and what the options do.

It's full of stupid
features like the "Tools", the "most powerful" of which, the template
tool, does the job of four lines of Perl or something similar.

What four lines of Perl can do the same thing? I don't think so. I'll be
very impressed if you can write the template tool in 4 typical Perl lines
(maybe you could if the lines are really, really, really long!). Anyway, the
tools are simple utilities that people find useful and not really the
subject of this discussion.

Now I don't say this product is completely useless to everyone. I tried
it on a couple of pages and the warning messages it produced were fairly
user-friendly and informative (although no more so than those produced
by tidy) and I even think the tools might be quite useful to a rank
novice.

Great, I'm glad you see some use.

But I'd rather novices learned about DTDs and real validators,
and how to use decent editors and how to think for themselves, because
none of that is really all that difficult if it's explained well, than
to get sidetracked by products like this.

Again, you're entitled to your opinion but I think CSE HTML Validator is an
excellent HTML/CSS learning tool. In fact, some instructors use it to teach
HTML to their students and find it very useful. Besides, DTDs and real
validators aren't that useful in practice, at least for HTML.

>As for the "parser bug", it's not a bug. It's designed to help people
>write better HTML by closing many of their optional tags.

>
Rubbish. If that were case it would have said "<dt>: warning end tag is
optional and you've left it off: I hope you know what you're doing" or
words to that effect.

Not rubbish. Just because it doesn't display your message doesn't mean what
I said was rubbish.

Albert

[Albert Wiersch]

Re: CSS software tools sought


"Chris F.A. Johnson" <cfajohnson@gma il.comwrote in message
news:ftcgk4-atj.ln1@206-248-139-163.dsl.teksavv y.com...

>
In other words, it is not a validator, since there is nothing wrong
with omitting the closing tags. It is a style checker.

It's not a DTD based validator, but it does include one.

Albert

[Chris F.A. Johnson]

Re: CSS software tools sought

On 2007-06-17, Albert Wiersch wrote:

>
"Chris F.A. Johnson" <cfajohnson@gma il.comwrote in message
news:ftcgk4-atj.ln1@206-248-139-163.dsl.teksavv y.com...

>>
> In other words, it is not a validator, since there is nothing wrong
> with omitting the closing tags. It is a style checker.

>
It's not a DTD based validator,

How else would you define valid?

The so-called errors that I've seen it produce have nothing to do
with validity.

but it does include one.

--
Chris F.A. Johnson <http://cfaj.freeshell. org>
========= Do not reply to the From: address; use Reply-To: ========
Author:
Shell Scripting Recipes: A Problem-Solution Approach (2005, Apress)